1. Field of the Invention
The invention relates to a storage system, a release method, and a secondary storage apparatus. The invention is particularly utilized in storage systems having a storage apparatus in which plural virtual volumes are paired.
2. Description of Related Art
In storage systems in which data is stored using storage apparatuses, there is a conventional method for managing plural hard disks in a RAID (Redundant Array of Independent/Inexpensive Disks) format. At least one logical volume is formed in a physical storage area provided by a number of hard disks.
In recent years, Japanese Patent Laid-Open Publication No. 2003-015915 has disclosed a technique of providing virtual volumes from logical volumes to a host computer, not creating logical volumes with a fixed capacity from a storage area in a hard disk. Storage areas in logical volumes are dynamically allocated to those virtual volumes. With this configuration, storage areas that are in actuality dispersed over plural storage apparatuses can be provided as a single volume to a host computer.
Two of virtual volumes are set as primary and secondary virtual volumes respectively and paired, and data reliability is maintained by copying data from the primary virtual volume to the secondary virtual volume.
If a new pair is set using a secondary virtual volume that has already been used, a method where all data (data including zero data) of the same size as the available capacity in the secondary virtual volume is copied from a relevant primary virtual volume to the secondary virtual volume to make the content of those virtual volumes the same, or a method where the secondary virtual volume is paired with a primary virtual volume after initializing the area that has been used in the secondary virtual volume and only a data storing area in the primary virtual volume is copied to the secondary virtual volume to make the content in those virtual volumes the same has been used.
However, since in the former method the amount of traffic is large, data transfer takes long time. Although in the latter method the traffic is small, a primary virtual volume has to be acquainted with the status of a relevant secondary virtual volume, and data copy has to be stored in a storage area in a secondary virtual volume corresponding to a data storing area in a primary virtual volume.
In addition, when a user pays an amount in proportion to the purchased capacity of both volumes set in a pair, in the case of the former method the user has to pay for not only the capacity necessary for data copy but also the capacity of a storage area that has already been used in a secondary virtual volume. In the case of the latter method, a user has to purchase only the capacity used for data copy, but the user has to initialize and thus reconfigure, by himself, a secondary virtual volume that has already been used for another purpose immediately before setting a pair. In this way, dealing with cost and operation redundancy becomes an issue for a user.